Small area models for unemployment rate estimation at sub-provincial areas in Italy (?)

Abstract: The goal of this paper is to analyze the possibility to improve the performance of the estimation at sub-regional level from ISTAT Labour Force Survey. In particular, we refer to estimation of unemployment rates for small domains cutting across survey strata, i.e. Local Labour Market Areas, defined as aggregation of municipalities. Currently, such quantities are estimated by means of EBLUP based on a linear mixed model with spatially correlated area effects and covariates given by sex, age and area level unemployment rate at previous census. In this work we investigate the use of nonparametric additive models based on penalized splines to prevent misspecifications of the functional relationship between the target variable and some covariates. Moreover, in the same way, nonparametric additive modelling can be used to include spatial information using low-rank thin plate splines (Opsomer et al, 2008). In addition, we analyze the possibility of reducing bias of EBLUP through the use of model based direct estimation (Chambers e Chandra, 2005). Finally, small area estimators based on logistic (mixed) models are explored to account for the binary nature of the response variable more appropriately. The performances of the aforementioned methods are studied via simulated experiments on 2001 Census data